The Sinister Side of the Internet: A Look into the Dark Future of A.I.

In October, the Louisiana parole board discussed the potential release of a convicted murderer and brought in a doctor with mental health experience to talk about the inmate. However, online trolls on 4chan took screenshots of the doctor from an online feed of her testimony, edited the images with A.I. tools to make her appear naked, and shared the manipulated files. Those tactics exemplified how users on 4chan use A.I. technology to promote racist and offensive content about people who have appeared before the parole board. Similarly, the article also highlights some problems resulting from the misuse of A.I. tools on 4chan and what regulators and technology companies are doing about them.

The article noted concerns about A.I. tools like Dall-E and Midjourney that generate fake pornography by removing clothes from existing images. Such manipulative images and audio, while not spreading far beyond 4chan, can be a cause of concern for efforts to supercharge online harassment and hate campaigns. The article also discussed the consequences of A.I.-generated pornography and the lack of federal law against its creation. Additionally, the potential for A.I. tools to be exploited for harmful purposes was exemplified by a tool that can create realistic digitally replicated voices and was used by users to circulate fake clips of celebrities expressing offensive content.

The article also mentioned how major social media companies and the government are taking steps to address the issue of A.I.-generated content. It was mentioned that some companies have begun labeling A.I. content, and President Biden issued an executive order for content labeling and development of standards for watermarking and authenticating A.I. content. The misuse of open-source A.I. tools was also highlighted, as Meta released its large language model, Llama, to select researchers, and the code quickly leaked onto 4chan. Users then used the code for various purposes, including producing antisemitic ideas. Moreover, language models have been developed to echo far-right talking points and create sexually explicit content.

Overall, the article exemplifies how fringe platforms like 4chan provide early indications of how new technology may be utilized for promoting extreme and harmful ideas and the steps being taken to address this misuse.

News

Moms Managing Girl Influencers: A Marketplace Stalked by Men

Elissa began receiving threatening messages early last year from a person calling themselves “Instamodelfan” targeting her daughter’s Instagram account. Despite having over 100,000 followers, the account has been under scrutiny for potentially exploiting children in exchange for money. However, the issue runs deeper than that. Research from The New York Times found that the platform […]

Read More
News

Insights from The Times’s Investigation of Child Influencers

Instagram maintains the 13-year-old minimum age for accounts, but parents can take control, largely for their daughters’ ambitions to become influencers. Parents initiate their child’s modeling career or gain favor from clothing brands, but a dark subculture emerges, controlled by men attracted to minors, as per The New York Times. The emergence of mom-run profiles […]

Read More
News

Rising Threat: China’s Growing Cyber Espionage and the New Vulnerability

Beijing’s Networks Expanding Hacking Efforts China has spread its hacking reach with new tools that exploit computer vulnerabilities and a network of contracted vendors. The large scale of China’s hacking operations poses a significant threat, with the FBI reporting China’s hacking program to be larger than all major nations combined. The U.S. has tracked consistent […]

Read More